Professional Documents
Culture Documents
Abstract: In biometrics, every individual must be identified based on some characteristic physiological parameters. A wide
variety of recognition schemes are used to either determine or confirm the identity of an individual requesting their services. In
the proposed work, different modules are combined to generate more secure cryptographic key from iris image. In the initial
stage, iris images from the CASIA Database V1.0 are processed to locate the iris region and this region is normalized into a
dimensionally constant rectangular block using Daugman's operator know as rubber sheet model. Principal component analysis
is used to match the rubber sheet models. Two unique prime numbers have been assigned for the selected portion of the rubber
sheet model which is to be utilized for the data encryption and decryption tasks. Most public-key cryptographic algorithms
require the generation of random prime numbers. A naive solution to obtain a prime number consists of randomly choosing a
number and testing it for primality. The generation of prime numbers underlies the use of most public-key cryptosystems,
essentially as a primitive needed for the creation of RSA key pairs. Accuracy of the data encryption and decryption algorithm
depends on the iris image segmentation. The algorithm was applied for 105 images of CASIA database and iris segmentation is
approximately 85.71% i.e. 90 out of 105 images are segmented successfully.
Keywords: Iris Recognition, RSA algorithm, Encryption, Decryption
I. INTRODUCTION
Use of Internet is growing rapidly. So providing security to the information over networks has become a crucial issue these days.
Information should be disclosed solely to the intended recipient. In the network, information is more open to attacks. Cryptography
techniques are used to prevent the sensitive information from the intruder. Encryption converts the original messages into coded
messages by making a system immune to different attacks. At present, the best well-known and generally used public key
cryptosystem for secure data transmission is RSA. In such a cryptosystem, the encryption key differs from the decryption key which
is kept secret. This asymmetric algorithm is based on the practical factoring of the product of two large prime numbers. These prime
numbers must be kept secret. Anyone can use the public key to encrypt a message but only with knowledge of the prime numbers
one can possibly decode the message.
Biometric methods that are based on the every individual verification and identification processes are gaining additional interest in
today’s era because of rapid growth in the technology and high requirement for security. Among different biometrics like
fingerprint, face recognition, iris and ear, iris recognition method is more accurate and reliable due to its unique statistical
characteristic and high identification rate. These systems can be divided into two stages, enrolment stage and verification stage. In
enrolment stage, eye images are captured using high quality camera and these eye images are pre-processed and the corresponding
iris templates are stored in the database. In verification stage, user asks to submit their samples first, these templates are compared
against the stored database and then decision is taken based on the threshold value. Basically a system is to be implemented to
provide security to data using iris image stored in CASIA database. There is an essential need for personal characteristics based
identification due to the fact that it can provide the highest protection against impersonation. This system consists of an image
segmentation which detects the circular iris region, occluding eyelashes and eyelids and reflections. Along with this RSA algorithm
is designed using iris template as key. The encryption and decryption processes work separately in combination with a key to
encrypt and decrypt the data. The encrypted data security totally depends on the confidentiality of the key.
concluded that confidentiality and scalability provided by Advanced Encryption Standard over Data Encryption Standard and Rivest
Shamir Adleman algorithm is much higher and makes it suitable.
Zhou et al. [3] discussed a most common encryption algorithm known as Rivest, Shamir and Adleman algorithm with the real fact
that it's very difficult to factorize big integers. It is the first asymmetric cryptosystem which can be used in both key exchange and
digital signatures. It has block size of 128 bits and key size varies from 1024 to 4096. Two different keys are utilized for encoding
and decoding the data. It makes use of two prime numbers to generate public and private keys based on mathematical calculations
and multiplying those large numbers together.
J.Daugman [6] designed and patented the first complete, commercially available phase-based iris recognition system in 1994. The
eye images with resolution of 80-130 pixels iris radius were captured with image focus assessment performed in real time. The iris
and pupil boundaries are detected using integro-differential operator. The centre coordinates and radius are estimated for both pupil
and iris regions by determining the maximum partial derivative of the contour integral of the image along the circular arc. The iris
portion of the image is normalized. Each pixel in the normalised iris pattern corresponds to two bits of data in the iris template. The
representation of iris texture is binary coded by quantizing the phase response of a texture filter using quadrature 2D Gabor wavelets
into four levels. This algorithm achieves high performance in iris recognition.
Masek [7] designed an open iris recognition system for the verification of human iris uniqueness and also its performance as the
biometrics. The iris recognition system consists of an automated segmentation system, in which iris region is isolated from the
original eye image using circular Hough transform.. By applying Daugman’s rubber sheet model the segmented iris region
converted into flexible rubber sheet rectangular block. Ultimately, the iris features were encoded by convolving the normalized iris
region with the 1D Log-Gabor filters and phase quantizing the output to produce a bit-wise biometric template.
i) If P = 11 and Q = 17
ii) Then N = P x Q
= 11 x 17
N = 187
iii) Compute Φ(N) = (P - 1) x (Q - 1)
= (11-1) x (17-1)
= 10 x 16
Φ(N) = 160
iv) Choose an integer E such that 1 < E < Φ(E) , E and Φ(N) are co-prime.
v) Let E = 13
vi) Determine D = E-1 mod (Φ(N))
= 13-1 mod (160)
D = 37
vii) Public key (E, N) = (13, 187)
viii) Private key (D, N) = (37, 187)
ix) Encryption : If M = 20 then ,
C = ME mod N
C = 2013 mod 187
C = 80
x) Decryption : C = 80 then,
M = CD mod N
M = 8037 mod 187
M = 20
Iris database
Image Segmentation
Image Normalization
Principal component
analysis
Matching
1) Image Segmentation: In iris recognition method, first step is to separate the actual iris region in original eye image. The iris
region is separated by two circles, one is the limbus or sclera boundary and another one is iris/pupil boundary. The upper and lower
parts of the iris region are normally covered by eyelids and eyelashes. A technique is required to isolate and eliminate these artifacts
as well as locating the circular iris region. An integro-differential operator is used for detecting the iris and pupil areas. The
Daugman's integro-differential operator is defined as,
( , )
max(r,xp,yo) =│Gσ (r) * ∮, ,
ds│ eq. (1)
where,
I(x,y) Eye image.
r Radius to search.
G (r) Gaussian smoothing function.
σ
s Contour of the circle given by r, x , y .
0 0
The operator creates a circular path by changing the radius r and centre x and y position if there is a maximum change in
the pixel values. This process is repeated in order to attain precise localization.
2) Image Normalization: After segmenting the iris part from an eye image, the second step is to convert this iris area into
rectangular form with fixed dimensions in order to allow comparisons. Irregularities in the dimensions are caused by the stretching
of the iris which leads to the errors in the calculations. This is done using Daugman's rubber sheet model, where each point in the
iris area is remapped into polar coordinates (r,θ) as shown in Fig. 2.
0 1 r
r
The iris region is remapped from Cartesian coordinates to normalized polar coordinates which is represented as,
I (x (r,ϴ), y (r,ϴ) ) I (r,ϴ) eq. (2)
with
x(r,ϴ) = (1- r) xp (ϴ) + rxl (ϴ)
y(r,ϴ) = (1- r) yp (ϴ) + ryl (ϴ)
where,
I(x,y) Iris region.
(x,y) Original Cartesian coordinates.
(r,θ) Normalized polar coordinates.
(xp ,yp ) & (xl , yl ) Coordinates of the pupil and iris boundaries along the θ direction.
3) Principal Component Analysis: Principal Components Analysis (PCA) is a data transformation technique invented by Karl
Pearson in 1901. This technique is very popular in the fields such as image compression and face recognition. If a number of
variables are to be measured for a series of objects or persons, then each variable will have a variance. These variables are related
each other i.e. there will be covariance between pair of variables. The data set used for recognition as a whole will have a total
variance which is the sum of the individual variances. 105 iris images from the CASIA database are selected, obtained two
dimensional iris template is converted into one dimensional template.
The steps involved in PCA are listed below :
IV. RESULTS
A. RSA Algorithm
RSA is a strong cryptographic technique which can be used to encrypt and decrypt any given data. The algorithm asks user to enter
the two prime numbers P & Q, then it calculates the value of N, Φ ,E & D. After generating the public and private keys it asks user
again to provide input data that is to be encrypted. The encryption and decryption results for a given data is as shown Fig. 3.
Linear Sieve algorithm is a fast method for finding all the prime numbers within a given range. It gives prime numbers by
repeatedly marking all the composite numbers which are the multiples of each prime, starting withѧ2. For a given maximum limit
60, the generated prime numbers are as shown in Fig. 5.
A prime number is twin prime if the difference between them is 2 i.e. a prime number is either 2 more or 2 less than the other prime
number. The twin prime generator algorithm generates all the twin primes within a given range. For the given value 50, the
generated prime numbers are as shown in Fig. 6.
Another simple method is presented for generating the prime numbers within a given maximum range. For the given range 40, the
generated prime numbers are as shown in Fig. 7.
(a) Original Image (b) Outer Edge Detection (c) Inner Edge Detection
(e) Left portion of rubber sheet model (f) Right portion of rubbers sheet model
The prime numbers P & Q are selected from the left part of rubber sheet model, then it calculates the values of N & Φ. After
generating the private and public keys it asks user to enter the message that is to be encrypted. This encrypted message is stored in
MATLAB file known as "encrypted_mesg" file as shown in Fig. 9.
In RSA decryption part, different eye image of a same person is selected from the CASIA database. Then the iris region
segmentation and normalization processes are carried out. Once the rubber sheet model is obtained, the left portion of it to which
prime numbers have been assigned is selected. These prime numbers are used as input to the decryption part.
The prime numbers used in this process should be same as that of the encryption process, then it calculates the values of N & Φ.
After generating the private and public keys it asks user to enter the file in which the encrypted message is stored. The decrypted
message is as shown in above Fig. 10 , which is same as the input message.
In this algorithm, two separate images of a same person are used for encrypting and decrypting of a given data. This method will be
successful only if the iris image segmentation and matching is done appropriately. Otherwise, for a given data encryption and
decryption processes will be failed. Segmentation is a main step in iris recognition, because the part which is wrongly identified as
iris will corrupt the iris templet resulting in poor recognition. So accuracy of the algorithm for segmentation is approximately
85.71% i.e. 90 out of 105 images are segmented successfully.
V. CONCLUSION
With the rapid growing of internet and networks applications, security of information becomes more important. Encryption
techniques play very important role in information security. In this project work, the performance of RSA algorithm is studied and
analysed. Prime numbers are generated using various algorithms and verified their primility property. To meet current security
requirement, iris recognition system is used to protect the keys from intruder. The performance of the iris recognition system is
verified, where an automatic iris region segmentation method is presented using Daugman's operator. This iris region is normalized
using Daugman's rubber sheet model, where the circular iris section is converted into rectangular block from which only 100 X 150
pixels are selected excluding eyelashes and eyelid part to avoid errors in the matching process. Finally, two unique prime numbers
have been assigned to the selected rubber sheet model which acts as input to the modified RSA algorithm. For the matching of
rubber sheet models PCA technique and Euclidean distance are used and their performance is verified. This method will be
successful only if the iris image segmentation and matching is done properly. Otherwise, for a given data encryption and decryption
processes will be failed. Segmentation is a main step in iris recognition, because the part which is wrongly identified as iris will
corrupt the iris templet resulting in poor recognition. The algorithm is applied for 105 images of CASIA database and the precision
for segmentation is approximately 85.71% i.e. 90 out of 105 images are segmented successfully.
REFERENCES
[1] William, Stallings. Cryptography and network security: principles and practices. Pearson Education India, 2006.
[2] Mahajan, Prerna, and Abhishek Sachdeva. "A study of encryption algorithms AES, DES and RSA for security." Global Journal of Computer Science and
Technology (2013)
[3] Zhou, Xin, and Xiaofei Tang. "Research and implementation of RSA algorithm for encryption and decryption." In Strategic Technology (IFOST), 2011 6th
International Forum on, vol. 2, pp. 1118-1121. IEEE, 2011.
[4] Konigsberg, Zvi Retchkiman. "Primes and Twin Primes Generator Algorithms." In Computational Engineering in Systems Applications, IMACS
Multiconference on, vol. 2, pp. 1-4. IEEE, 2006.
[5] Pittu, Ganesh Reddy. "Generating primes using partitions." arXiv preprint arXiv:1505.00253 (2015).
[6] Daugman, John G. "High confidence visual recognition of persons by a test of statistical independence." IEEE transactions on pattern analysis and machine
intelligence 15, no. 11 (1993): 1148-1161.
[7] Masek, Libor. "Recognition of human iris patterns for biometric identification." (2003): 1-7.
[8] Wildes, Richard P. "Iris recognition: an emerging biometric technology." Proceedings of the IEEE 85, no. 9 (1997): 1348-1363.
[9] Krichen, Emine, M. Anouar Mellakh, Sonia Garcia-Salicetti, and Bernadette Dorizzi. "Iris identification using wavelet packets." In Pattern Recognition, 2004.
ICPR 2004. Proceedings of the 17th International Conference on, vol. 4, pp. 335-338. IEEE, 2004.
[10] Huang, Ya-Ping, Si-Wei Luo, and En-Yi Chen. "An efficient iris recognition system." In Machine Learning and Cybernetics, 2002. Proceedings. 2002
International Conference on, vol. 1, pp. 450-454. IEEE, 2002.
[11] Jong Gook Ko, Youn Hee Gil, Jang Hee Yoo, and Kyo Il Chung. "Method of iris recognition using cumulative-sum-based change point analysis and apparatus
using the same." U.S. Patent 7,715,594, issued May 11, 2010.
[12] Zhou, Xin, and Xiaofei Tang. "Research and implementation of RSA algorithm for encryption and decryption." In Strategic Technology (IFOST), 2011 6th
International Forum on, vol. 2, pp. 1118-1121. IEEE, 2011.
[13] Gries, David, and Jayadev Misra. "A linear sieve algorithm for finding prime numbers." Communications of the ACM 21, no. 12 (1978): 999-1003.
[14] Marc Joye, Pascal Paillier, and Serge Vaudenay. "Efficient Generation of Prime Numbers." In Cryptographic Hardware and Embedded Systems-CHES 2000:
Second International Workshop Worcester, MA, USA, August 17-18, 2000 Proceedings, p. 340. Springer, 2003.
[15] Buddharpawar, Aniket S., and S. Subbaraman. "Iris Recognition based on PCA for Person Identification." International Journal of Computer Applications
(0975–8887)(2015).
[16] Shi, Jin-Xin, and Xiao-Feng Gu. "The comparison of iris recognition using principal component analysis, independent component analysis and Gabor
wavelets." In Computer Science and Information Technology (ICCSIT), 2010 3rd IEEE International Conference on, vol. 1, pp. 61-64. IEEE, 2010.
[17] Jagadeesh, N., and Chandrasekhar M. Patil. "Iris recognition system development using MATLAB." In Computing Methodologies and Communication
(ICCMC), 2017 International Conference on, pp. 348-353. IEEE, 2017.